Abstract

We propose Hemera Protocol, a decentralized programmable indexing network that enables developers to harness web3 data with ease and unlock the full potential of the ever-growing network of data in web3. Hemera is missioned to “accelerate the advent of data and AI democratization”, and achieves so through building a public goods infrastructure that is decentralized in nature and owned by the community. Hemera tackles the data interoperability challenges by solving the basic and yet expensive data retrievability needs through an extensible architecture. Numerous immediate web3 use cases, ranked from simple to complex, including blockchain explorers, on-chain assets portfolio, social graph, zero-knowledge (ZK) coprocessors, community quality auditing, web3 identities, etc., can thus benefit from or built on top of Hemera. Future compatibility with Large Language Models (LLMs) and broader AI models are built into the architecture of Hemera from day one, with deeper synergies being actively explored and briefly discussed here.

Ever since Ethereum was conceived back in 2013 [1] as a fully programmable distributed supercomputer, various protocols have been built to bring user activities on chain, ranging from the more well-known financial activities, to everyday activities including Non-Fungible Token (NFT) communities, gaming, social, Decentralized Autonomous Organizations (DAOs) etc. The network of data composed of records of activities is promised to empowe web3 developers to build application experiences beating web2 counterparts.